Estuarine, Coastal and Shelf Science 86 (2010) 331–336
Contents lists available at ScienceDirect
Estuarine, Coastal and Shelf Science
journal homepage: www.elsevier.com/locate/ecss
Study on the total water pollutant load allocation in the Changjiang
(Yangtze River) Estuary and adjacent seawater area
Yixiang Deng *, Zheng Binghui, Fu Guo, Lei Kun, Li Zicheng
Chinese Research Academy of Environmental Sciences, Beijing 100012, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 23 February 2009
Accepted 25 October 2009
Available online 18 November 2009
With the rapid economic development, the water quality is worsening and red tide takes place frequently in
the Changjiang Estuary and adjacent seawaters. To improve the marine water quality, the total inland
pollutant load should be controlled effectively. With efficiency and fairness in consideration, the total
maximum allowable loads of CODMn, NH3–N, inorganic nitrogen and active phosphate to the seawaters were
calculated and allocated by the linear programming method based on the water quality response fields of
the pollution sources. The maximum allowable loads are 2008 103 tons, 169 103 tons, 226 103 tons and
18 103 tons for CODMn, NH3–N, inorganic nitrogen and active phosphate when the water quality targets are
requested to be achieved in the whole studied region, and 346 103 tons and 32 103 tons for inorganic
nitrogen and active phosphate when the water quality targets to be achieved only in the red tide sensitive
area. The cut task of CODMn and NH3–N is relatively easy and can be finished by the watershed environmental plan; while the cut task of inorganic nitrogen and active phosphate is tremendous. The coastal
provinces should install more denitrification and dephosphorization facilities in the existing waste water
treatment plants or build new ones to control the red tides in the concerned seawaters.
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Changjiang Estuary and adjacent seawaters
total pollutant load allocation
denitrification
dephosphorization
pollution control
1. Introduction
In the recent 30 years, the economy has been developing rapidly
in the Changjiang (Yangtze River) Watershed, especially in the
Changjiang Delta area. The Changjiang Estuary and adjacent
seawaters become one of the most seriously polluted marine areas
in China due to the accelerated urbanization and industry development (Wang, 2006; Chen et al., 2007; Zhang et al., 2007; Zhou
et al., 2007; Lei et al., 2009). From 2005a to 2006a, the Ministry of
Environment Protection (MEP) organized a water quality investigation in this area. According to the monitoring data, the ratios of
the surface water areas with the first and second grade of CODMn
water quality standards were 93.0% and 6.9%; the ratios of the
surface water areas with the forth and inferior forth grade of active
phosphate standards were 33.6% and 14.5%, and the ratio of the
areas with the inferior forth grade of inorganic nitrogen standard
was 62.6%. The inorganic nitrogen and active phosphate were the
primary pollutants.
In order to improve the water quality and ecological environment, the total pollutant load from inland to the seawaters must be
controlled. Based on the response relationship between the
* Corresponding author.
E-mail address: [email protected] (Y. Deng).
0272-7714/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ecss.2009.10.024
pollutant load and marine water quality, the total maximum
allowable pollutant load was calculated and allocated to the
pollution sources for the coastal provinces and upstream region.
2. Study site
The Changjiang River is the longest river in China. It lies in the
middle of China and flows from west to east until it impounds into
the China East Sea. The Changjiang Estuary and adjacent seawaters
(including Qiantangjiang River Estuary) are halfly surrounded by
two provinces and one city, i.e. Jiangsu, Zhejiang, and Shanghai.
Based on the marine function zoning programme enacted by the
State Environmental Protection Administration (SEPA, former MEP)
in 2001 (SEPA, 2002), the water quality standard objectives were
set and some modifications were made according to the local water
quality standards in Shanghai City and Jiangsu Province. In the
inner part of the Changjiang Estuary, the water quality objectives
were set by the Environmental Quality Standards for Surface Water
(GB3838-2002, enacted in 2002) (Fig. 1).
The thousands of pollution sources in the research area were
grouped into 33 conceptual pollution sources to avoid extreme
complexity and confusion, including the major rivers such as
Changjiang River, Qiantangjiang River, Huangpujiang River,
Chaoejing River, Yongjiang River and Jiaojiang River, and 2 direct
seawater discharge pollution sources in Jiangsu, 8 in Shanghai and
332
Y. Deng et al. / Estuarine, Coastal and Shelf Science 86 (2010) 331–336
Fig. 1. Marine water quality control objectives in the research area.
17 in Zhejiang (Fig. 2). Totally 2245 103 tons CODMn, 194 103
tons NH3–N, 136 103 tons inorganic nitrogen and 82 103 tons
active phosphate were discharged into the Changjiang Estuary and
adjacent seawaters in 2005. The marine aquaculture pollution and
atmospheric deposition were treated as environmental background
and not considered in the total load allocation in this study.
Two types of pollutants were considered. The first was the
organic pollutants, including CODMn and NH3–N. The second was
the eutrophication pollutants, including dissolved inorganic
nitrogen and inorganic phosphate. The Marine Water Quality
Standard (GB 3097-1997) was adopted to set the water quality
target for the above pollutants.
The design flow condition was also set differently for the above
two types of pollutants. The lowest monthly average flow was
selected as the design flow condition for CODMn and NH3–N to
ensure that the water quality standards can be met in most of the
year. The annual average flow of multiple years was selected as
design flow condition for inorganic nitrogen and active phosphate
as the eutrophication is closely relative to the substance accumulation in a long time rather than the worst hydraulic condition.
The total loads of all pollutants, i.e. CODMn, NH3–N, inorganic
nitrogen and active phosphate, were firstly allocated under the
condition that the water quality targets were to be met in whole
research region. Then the total loads of inorganic nitrogen and active
phosphate were allocated under the condition that the water quality
targets were only to be met in the red tide sensitive area. The red tide
sensitive location was determined by Qi’s research (Qi, 2003).
3. Method description
The linear programming method based on response field of
pollution source was used in the total pollutant load allocation.
Firstly the response field of each pollution source was calculated
through water hydraulic and quality modeling, and the response
Y. Deng et al. / Estuarine, Coastal and Shelf Science 86 (2010) 331–336
333
Fig. 2. Location of the conceptual pollution sources (river included). *The rectangle area is the red tide sensitive area in the research area.
relationship between the pollutant emission and water quality was
built up; secondly the optimization target function and water
quality constraint equations were formulated, and the linear
programming method was used to calculate the environmental
capacity; lastly the total maximum allowable load was allocated
with the fairness among the pollution sources in consideration. The
water quality model, linear programming model and total pollutant
load allocation model were used respectively in the three steps.
3.1. Water quality model
The MIKE21 model was used in this research. The basic pollutant
transportation equation is convection–diffusion equation (Barettaa
et al., 1994; Chubarenko and Tchepikova, 2001; Lin et al., 2008):
vC=vt þ u vC=vx þ v vC=vy ¼ Dx v2 C=vx2 þ Dy v2 C=vy2 þ s þ KC
where C – pollutant concentration (mg/L); t – time (s); x,y – the
latitudinal and longitudinal coordination (m); u,v – velocity
component in x,y direction (m/s); Dx,Dy – turbulent diffusion
coefficient in x,y direction (m2/s); K – comprehensive degradation
coefficient; s – source and sink item.
The response field of each pollution source was calculated with
the above water quality model under the load with concentration of
1.0 mg/L of CODMn, NH3–N, inorganic nitrogen and active phosphate respectively.
3.2. Linear programming model
The maximum environmental capacity was calculated only
considering the environmental dilution and purification ability. The
fairness among the pollution sources was not considered in this
step. The formulas of the linear programming used in the capacity
optimization calculation are (Appan, 1989; Zhang, 2007):
334
Y. Deng et al. / Estuarine, Coastal and Shelf Science 86 (2010) 331–336
T
max
8z ¼ C X
< AX þ B S
st: Xl X Xu
:
X0
4. Results
(1)
where, z – objective function, C – coefficient vector. When the total
pollutant load is considered, set C ¼ [1, 1, ., 1]T.
A is the response matrix:
2
a11
A ¼ 4 «
am1
/
aij
/
3
a12
« 5
amn
(2)
The 2-dimensional MIKE21 model was used to build the relationship matrix between the pollutant load and the water quality
response. To overcome the effects of the errors in the initial field,
the stable calculation time duration was 2 years. More than 5 h
were needed to calculate the response field for each pollution
source on the computer. So the nonlinear optimization method was
infeasible in the total pollutant load allocation in the Changjiang
Estuary and adjacent seawaters. The linear programming method
based on the response field was used in this study. The water
where aij – response concentration of the jth pollution source on
the ith water quality control point, which was calculated by MIKE21
model in this study; m – number of the water quality control
points; n – number of the pollution sources.
B is the background concentration vector, B ¼ [b1, b2, ., bm]T. S is
the water quality standard vector, S ¼ [s1, s2, ., sm]T. X, Xl and Xu are
the pollutant load, the upper and lower load limit vectors, X ¼ [x1,
x2, ., xn]T, Xl ¼ [xl1, xl2, ., xln]T, Xu ¼ [xu1, xu2, ., xun]T
3.3. Total pollutant load allocation model
To make the allocation more acceptable for the polluters, the
principles of fairness and efficiency were considered at the same
time on the basis of the environmental capacity estimation. 6
factors were considered in this step, i.e. environmental capacity,
water resource, population, agriculture and economic development. The ratio of the allocated load of each pollution source should
be close to its ratios of the above six factors as much as possible. So
the total pollutant load allocation was a multiple objective problem.
The weighted average method was used to calculate the share ratio
of each pollution source (Fu et al., 2006; Meng, 2008):
ri ¼
6
X
wk rki
(3)
k¼1
where, ri is the share ratio of allocated load of the ith pollution source;
w1, w2, ., w6 are the weights for the environmental capacity, existing
discharge load, water resource, population, agriculture land area and
P6
wi 0ði ¼ 1; 2; .; 6Þ; r1i,
economic added value,
i ¼ 1 wi ¼ 1;
r2i, ., r6i are the share ratios of the 6 factors for the ith pollution
source.
In this study, the weights of the 6 factors were 0.421, 0.079,
0.184, 0.103, 0.122, and 0.092 based on the more than 20 investigation tables in the two provinces and one city.
Set R as the ratio vector of the pollution sources, R ¼ [r1, r2, .,
rn]. The pollution source load can be expressed as:
Y ¼ aR
(4)
where, a – the scalar parameter, Y – pollution load vector, Y ¼ [y1, y2,
., yn]. Then the vector Y can be replaced into the formula (1):
aAR þ B S
(5)
The value of a can be obtained as:
a ¼ min ðsi bi Þ=ðARÞi
1im
(6)
So the load estimation of ith pollution source is:
yi ¼ ri min ðsi bi Þ=ðARÞi
1im
(7)
Herein the allocation load of each pollution source can be solved
directly.
Fig. 3. Response fields of the pollution sources (mg/L). (a) COD for Changjiang River.
(b) COD for Qiangtiangjiang River.
Y. Deng et al. / Estuarine, Coastal and Shelf Science 86 (2010) 331–336
a 3000
b 160
140
Present load
Allocated load
2000
1500
1000
NH3-N(1000t/a)
CODMn(1000t/a)
2500
500
Present load
Allocated load
120
100
80
60
40
20
0
0
Jiangsu
Shanghai
c 1400
Present load
Allocated load 1
1200
1000
Changjiang
Zhejiang
Allocated load 2
800
600
400
d
DIP(1000t/a)
Changjiang
DIN(1000t/a)
335
Jiangsu
Shanghai
Zhejiang
40
30
Present load
Allocated load 1
25
Allocated load 2
35
20
15
10
200
5
0
0
Changjiang
Jiangsu
Shanghai
Zhejiang
Changjiang
Jiangsu
Shanghai
Zhejiang
Fig. 4. Comparison of the allocated loads and the present loads in 2005. *Allocated load 1 means the loads were allocated to meet the water quality targets in the whole research region;
allocated load 2 means the loads were allocated to meet the water quality targets only in the red tide sensitive area. (a) CODMn. (b) NH3–N. (c) Inorganic nitrogen. (d) Active phosphate.
quality of the research field was the summary of the individual
response fields of all the pollution sources. So the complexity of the
calculation was greatly simplified. The CODMn response fields of
Changjiang River and Qiantangjiang River see Fig. 3.
With the above method, the total load was allocated among the
33 pollution sources (Fig. 4). The maximum allowable loads are
2008 103 tons, 169 103 tons, 226 103 tons and 18 103 tons
for CODMn, NH3–N, inorganic nitrogen and active phosphate when
the water quality targets are requested to be met in the whole
region, and 346 103 tons and 32 103 tons for inorganic nitrogen
and active phosphate when the water quality targets to be met only
the red tide sensitive area.
The cut rates of the pollutants grouped by region see Table 1.
For the Changjiang upstream region, the cut rates for CODMn and
NH3–N are relatively low, but the cut task for inorganic nitrogen
and active phosphate is still heavy. The control emphasis for the
upstream provinces is inorganic nitrogen and phosphate removal.
The status of Jiangsu province is similar to the Changjiang upstream
region. The emphasis is inorganic nitrogen and active phosphate
removal as well. For Shanghai, besides the cut task of inorganic
nitrogen and active phosphate is heavy, the CODMn and NH3–N cut
task is not easy either. For Zhejiang, the cut task for NH3–N, inorganic nitrogen and active phosphate is prominent.
5. Discussion
listed in the watershed programme. For the pollution sources with
huge direct discharge, the COD and NH3–N should also be
controlled under mixing zone constraints.
The inorganic nitrogen and active phosphate are the major
control pollutants. As the nitrogen and phosphate are mainly from
agriculture, the treatment emphasis should be paid more on nonpoint sources, such as reduction of fertilizer usage gradually,
establishment of ecological agriculture, and enhancement of
Natural Reservation Area protection, artificial wetland protection
and control of water and soil loss. The urban sewage treatment
plants should be equipped with nitrogen and active phosphate
removal facilities, or new denitrification and dephosphorization
waste water treatment plants must be built in the future.
It is very interesting to find that the active phosphate concentration does not increase obviously when the history data of 1959
and 1957 was compared with the observation data in 2005 and
2006. It is probably due to that although the phosphate load from
the inland is increasing, the marine fishery production has been
growing at the same time in these years, so more phosphate is
absorbed and the new balance is built up.
In this sense, the nitrogen removal is even more urgent than
phosphate. So recently the emphasis will be focused on nitrogen
removal. The active phosphate should also be controlled to prevent
the red tide comprehensively. As the nitrogen load removal task is
tremendous especially in the Changjiang upstream region, it will
take a long-term to realize the expected targets of nitrogen control.
The allocation results show that the cut task for COD and NH3–N
is relatively low. The control countermeasures are suggested to be
Acknowledgements
Table 1
Cut rates of the pollutant loads (based on 2005 data) to meet the water quality
targets by region.
The research was supported by MEP project of BLUE SEA ACTION
PLAN IN CHANGJIANG ESTUARY AND ADJACENT SEAWATERS.
Region
CODMn NH3–N Inorganic nitrogen
Active phosphate
Red tide
Whole
Red tide
Whole
controla sensitive
controla sensitive
b
area control
area controlb
Changjiang
Jiangsu
Shanghai
Zhejiang
Total
a
b
26.6%
27.5%
47.1%
47.3%
29.2%
5.7%
12.1%
85.2%
86.0%
39.7%
83.4%
83.5%
91.2%
83.1%
83.9%
76.0%
75.7%
82.5%
65.9%
75.4%
54.5%
58.5%
88.8%
57.5%
57.1%
26.3%
28.4%
44.6%
0.0%
23.8%
To meet the water quality standards in the whole research area.
To meet the water quality standards only in the red tide sensitive area in Fig. 2.
References
Appan, A., 1989. Water quality management in unprotected catchments – the total
approach. Environmental Software 4, 168–178.
Barettaa, J.W., Ruardijb, P., Vestedc, H.J., 1994. Eutrophication modelling of the North
Sea: two different approaches. Ecological Modelling 75/76, 471–483.
Chen, X.H., Zhu, L.S., Zhang, H.S., 2007. Numerical simulation of summer circulation
in the East China Sea and its application in estimating the sources of red tides in
the Yangtze River estuary and adjacent sea areas. Journal of Hydrodynamics 19,
272–281.
Chubarenko, I., Tchepikova, I., 2001. Modelling of man-made contribution to salinity
increase into the Vistula Lagoon (Baltic Sea). Ecological Modelling 138, 87–100.
336
Y. Deng et al. / Estuarine, Coastal and Shelf Science 86 (2010) 331–336
Fu, G., Deng, Y.X., et al., 2006. Report on the Technical Specifications of Water
Environmental Capacity Calculation. Chinese Research Academy of Environmental Sciences, Beijing, 512 pp. (in Chinese).
Lei, G., Dao, J.L., Ping, X.D., 2009. Quasi-simultaneous observation of currents,
salinity and nutrients in the Changjiang (Yangtze River) plume on the tidal
timescale. Journal of Marine Systems 75, 265–279.
Lin, W.Q., Lu, S.Q., Jiao, J.Z., 2008. Water quality and ecological modelling for
Yangtze Estuary and its adjacent sea. Chinese Journal of Hydrodynamics 23,
522–531.
Meng, W., 2008. Technique of Total Amount Control for Water Pollutants in
Watershed and its Application. Chinese Environmental Science Press, Beijing,
263 pp. (in Chinese).
Qi, Y.Z., 2003. China Coastal Red Tide. Science Press, Beijing, 326 pp. (in Chinese).
SEPA, 2002. Environmental Function Zoning in the China Coastal Seawaters.
Chinese Environmental Science Press, Beijing, 162 pp. (in Chinese).
Wang, L.D., 2006. Changjiang ecology and environment – present situation and challenges. Resources and Environment in the Yangzte Basin 15, 614–618. in Chinese.
Zhang, G.Z., 2007. Linear Programming. Wuhan University Press, Wuhan, 370 pp.
(in Chinese).
Zhang, J., Liu, S.M., Ren, J.L., Wu, Y., Zhang, G.L., 2007. Nutrient gradients from the
eutrophic Changjiang (Yangtze River) Estuary to the oligotrophic Kuroshio
waters and re-evaluation of budgets for the East China Sea Shelf. Progress in
Oceanography 74, 449–478.
Zhou, M.J., She, Z.L., Yu, R.C., 2007. Coastal ecosystem responses to Changjiang
nutrient inputs from large temperate and subtropical rivers. Continental Shelf
Research 28, 1483–1489.